Developing Clean Fuels: Novel Techniques for Desulfurization James P. Nehlsen A Dissertation Presented to the Faculty of Princeton University in Candidacy for the Degree of Doctor of Philosophy Recommended for Acceptance by the Department of Chemical Engineering November 2005
145
Embed
Developing Clean Fuels: Novel Techniques for Desulfurization
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Developing Clean Fuels: Novel Techniques for Desulfurization
James P. Nehlsen
A Dissertation Presented to the Faculty of Princeton University in Candidacy for the Degree of Doctor of Philosophy
Recommended for Acceptance by the Department of Chemical Engineering
Table 2.1: Petroleum distillate fractions and their boiling points. ....................................16
Table 2.2: Aromatic sulfur species found in petroleum. ...................................................17
Table 3.1: Relative rates of hydrogenation of olefin and ketone species in a PEMHR. ...42
Table 3.2: Fitting parameters for all mechanisms. Current density, i, has units mA cm-2 except in the power laws in which the units are nmol sec-1 cm-2. .............................48
Table 4.1: Predicted heats of reaction of metal oxides with observed reactivity. .............83
Table 4.2: Thiol content of several light streams used in gasoline production. ................88
Table 5.1: Partition coefficients for the sulfides studied. ................................................127
Figures
Figure 1.1: Forest in Poland destroyed by sulfur emissions from a coal-fired power plant without sulfur emission controls. Courtesy of the US Geological Survey. ...............11
Figure 2.1: Example recombination reaction producing an alkanethiol in an HDS reactor....................................................................................................................................18
Figure 2.2: A conventional slurry hydrogenation reactor and associated processes. Adapted from 41. ........................................................................................................25
Figure 3.1: Photographs of sample MEAs created by the airbrushing technique. On left, an MEA after pressing showing good dispersion of Ag powder over the initially black carbon cloth. On right, spraying of Cu powder onto carbon cloth using the airbrush. The carbon cloth is pinned at one edge between glass slides during spraying. .....................................................................................................................35
Figure 3.2: Photgraphs of the membrane reactor. Clockwise from top left are 1) the two reactor plates with gaskets; 2) detail of one reactor plate showing flow channels and o-ring groove; 3) previous version of reactor in operation with galvanostat and cartridge heaters attached; 4) previous version of reactor showing flow channels and gaskets; and 5) assembled reactor showing aluminum backing plates and Teflon tubing adapters. ..........................................................................................................37
Figure 3.4: Rate of hydrogenation of 1-decene over various catalysts. Lines are fits to ki0.5. The values of k are listed in Table 3.2. .............................................................40
Figure 3.5: Rate of hydrogenation of 1-decene over ineffective catalysts. Data enlarged from Figure 3.4, note the change in scale. Uncertainty in the data is considerable, shown by sample error bars for Cu. Line is an example fit to i0.5. ............................41
Figure 3.6: Rate of hydrogenation of acetone over Pt and Cu catalysts. Lines are fits to the Tafel / Langmuir-Hinshelwood mechanism and the power law, ki0.75. The fit parameters are listed in Table 3.2. .............................................................................43
Figure 3.7: Comparison of two common hydrogenation mechanisms. Rate equations are simplified by assuming constant surface coverage of organic. F is Faraday’s constant, U is the unsaturated organic species, and S is the saturated organic species....................................................................................................................................46
Figure 3.8: Rate of hydrogenation of 1-decene over Pt and Pd catalysts (from Figure 3.4) compared to Langmuir-Hinshelwood and Eley-Rideal style mechanisms. The fit of the power law ki0.5 is also shown. ..............................................................................49
Figure 3.9: Reaction mechanism for the dehydrogenation of isopropanol and electrolysis of water. The removal of the first electron is rate limiting in each mechanism. .......56
Figure 3.10: Rate of dehydrogenation of isopropanol over Pt and C anodes. Lines are fits to EQ. Fitting parameters for Pt are kw/kiso = 0.0135, αw-αiso = 0. Parameters for C are kw/kiso = 0.0102, αw-αiso = 0.026. ........................................................................59
Figure 3.11: Cyclic voltammogram of an Ag electrode in a tetrahydrothiophene solution. The anodic (oxidation) current increases with sulfur concentration, indicating the production of more Ag2S. The smaller cathodic wave near 0 V is the reduction of some of the Ag2S. ......................................................................................................64
Figure 4.1: TGA results showing decomposition of Pb(SC8H17)2.....................................72
Figure 4.2: WAXD pattern of recrystallized Pb(SC8H17)2. Calculated average d-spacing is 26.2 Å. .......................................................................................................................73
Figure 4.3: Sketch of likely structure of a Pb(SC8H17)2 crystal.........................................74
Figure 4.4: Graphical representation of predicted heats of formation of metal thiolates from the oxide. ...........................................................................................................81
Figure 4.5: Graphical representation of observed metal oxide activity. ............................82
Figure 4.6: The Exomer process, designed by ExxonMobil and Merichem, Inc. to remove high molecular weight recombinant mercaptans. Adapted from 97 ...........................89
viii
Figure 4.7: Diagram of the thiol removal process using lead oxides, showing the key operations including reaction, separation, and extraction. ........................................92
Figure 4.8: The yield of thiolates, indicating the degree to which PbO was converted into Pb(SR)2 in excess thiol. .............................................................................................96
Figure 4.9: Reaction rate of PbO and octanethiol with water added and water removed from the reactor........................................................................................................104
Figure 4.10: Batch reaction rate of PbO and octanethiol. Solid line is fit of the rate law shown to the data using a two parameter fit, with parameter values given in the figure. .......................................................................................................................105
Figure 5.1: ASR-2 oxidative desulfurization process developed by UniPure. Adapted from 37......................................................................................................................112
Figure 5.2: Photographs of the Tyndall sunset effect in a solution of toluene, tetrahydrothiophene, and thiophene.........................................................................118
Figure 5.3: GC/MS spectra for the reaction products of tetrahydrothiophene and sulfuric acid. The mass spectrum for product B matches that of tetrohydrothiophene 1-oxide. The spectrum of product A suggests a species with two oxygen atoms, such as the sulfone or ketosulfoxide. Mass spectrum C could be the adduct of sulfuric acid and tetrahydrothiophene. ................................................................................................120
Figure 5.4: Reaction rate of 2000 ppm solutions of thiophene, tetrahydrothiophene, and dibutyl sulfide with 16M H2SO4. Data are fit by a rate law that is first order in sulfide concentration, shown above. For the given rate constants, sulfide concentration is given in units of ppm S. ................................................................121
Figure 5.5: Consumption of tetrahydrothiophene from a 2000 ppm S solution after 1 hour reaction with various concentrations of sulfuric acid. Data is fit by a rate law first order in acid concentration shown. [H2SO4
*] is acid concentration above the threshold value shown by the intercept (2.4 M). .....................................................122
Figure 5.6: Reaction rate of tetrahydrothiophene from a 2000 ppm S solution reacting with 9M H2SO4. Rate is constant at 0.0055 min-1. ..................................................123
ix
List of Publications
Desulfurization Technologies:
1. Nehlsen, J. P.; Benziger, J. B.; Kevrekidis, I. G., Removal of alkanethiols from a hydrocarbon mixture by a heterogeneous reaction with metal oxides. Industrial & Engineering Chemistry Research 2003, 42, (26), 6919-6923.
2. Nehlsen, J. P.; Benziger, J. B.; Kevrekidis, I. G., A process for the removal of thiols
from a hydrocarbon stream by a heterogeneous reaction with lead oxide. Energy & Fuels 2004, 18, (3), 721-726.
3. Nehlsen, J. P.; Benziger, J. B.; Kevrekidis, I. G., Oxidation of Aliphatic and
Aromatic Sulfides Using Sulfuric Acid. Submitted to Industrial & Engineering Chemistry Research, July 2005.
Fuel Cell Design and Control:
5. Benziger, J.B., Nehlsen, J.P., Blackwell, D., Brennan, T., and Itescu, J., Water Flow in the Gas Diffusion Layer of Fuel Cells. In press, Journal of Membrane Science, Mar 2005.
6. Benziger, J.B., Satterfield, M.B., Hogarth, W.H.J., Nehlsen, J.P., and Kevrekidis, I.G., The Power Performance Curve for Engineering Analysis of Fuel Cells. In press, Journal of Power Sources, May, 2005.
7. Hogarth W.H.J., Nehlsen, J.P., and Benziger, J.B., Fuel Cell Design: The Impact of
PEM Fuel Cell Dynamics on Control and Systems Design. Submitted to AIChE Journal, May 2005.
8. Benziger, J.B., Hogarth, W.H.J., and Nehlsen, J.P., Fuel Cell Power and Efficiency: Universal Constraints. Submitted to IEEE Transactions on Power Conversion, July 2005.
Chapter 1 Introduction
10
1. Introduction
Energy production is one of the most pressing issues of modern times. Economic
activity and energy usage are intimately linked. The production of useful goods and
services requires energy, and more global economic output requires more energy usage.
World energy usage increased by an average of 1.7% annually from 1980-2001, to a total
of 404 quadrillion BTUs.1 Although the percentage of energy obtained from fossil fuels
declined over the same period, the share of world energy from fossil fuels is still over
82%, half of which comes from petroleum.1
Unfortunately, the predominant modern technique for producing energy, the burning of
fossil fuels, has a severe impact on the global environment. Some of this impact is the
result of impure fuels. Naturally occurring sulfur compounds left in fuels lead to the
emission of sulfur oxide gases. These gases react with water in the atmosphere to form
sulfates and acid rain which damages buildings, destroys automotive paint finishes, and
acidifies soil, ultimately leading to loss of forests and other ecosystems.2 Figure 1.1
illustrates the devastating effects unchecked sulfur emissions can have on the local
environment. Sulfur emissions also cause respiratory illnesses, aggravate heart disease,
trigger asthma, and contribute to formation of atmospheric particulates.3
Federal programs designed to reduce sulfur emissions from electric utilities and other
industrial sources have been successful. A cap-and trade program instituted by the EPA
in 1990 has led to decreased acidification of lakes and streams and an estimated human
health benefit of $70 billion. The cost of this program is estimated between $1-2 billion.2
Chapter 1 Introduction
11
Figure 1.1: Forest in Poland destroyed by sulfur emissions from a coal-fired power plant wi thout
sulfur emission controls. Courtesy of the US Geological Survey.
Chapter 1 Introduction
12
Utilities are not the only source for atmospheric sulfur. Automobiles are also adversely
affected by sulfur compounds. Sulfur levels in automotive fuels have a profound effect on
the efficacy of catalytic converters. Sulfur affects these emission control devices by
strongly adsorbing to the precious metal catalysts, preventing the adsorption and reaction
of hydrocarbons, nitrogen oxides, and carbon monoxide. The EPA estimates4 that
reducing sulfur levels from 400 ppm to 50 ppm reduces emissions of hydrocarbons by
45.9%, NOx by 7.01%, and CO by 31.12% (based on Tier 1 running specification) by
reducing the poisoning effect of sulfur. Obviously, emissions of SOx are also reduced by
an amount equivalent to the sulfur reduction. The US national average sulfur level in
automotive fuel in 1997 was 339 ppm.4
Producing energy in a clean and responsible manner can be accomplished in a number
of ways. The use of non-fossil fuel energy sources such as solar, wind, and nuclear power
will eventually replace fossil fuels. However, many of these technologies will require
many years before they are able to provide the amounts of energy needed. In the
immediate future, fossil fuel-based energy production will continue, and new
technologies need to be developed in order to produce clean fuels to power our societies.
The present work focuses on new ways to remove sulfur compounds from petroleum.
Sulfur compounds represent one of the most prevalent impurities found in crude oil. As
the world market for crude oil tightens with increasing demand, lower quality oils
containing higher levels of sulfur are used. Developing techniques to remove the sulfur
Chapter 1 Introduction
13
from these lower quality feedstocks efficiently will ensure that energy is available at a
reasonable cost.
The development of the new desulfurization techniques in this Thesis begins with a
study of polymer membrane reactors. These reactors offer a new way to perform
hydrogenation reactions including, possibly, hydrodesulfurization. The fundamental
reaction mechanisms in such reactors are studied to provide a basis for the development
of desulfurization techniques. Although polymer membrane reactors ultimately proved
unsuccessful for desulfurization, two new techniques for desulfurization are demonstrated
and evaluated.
First, the removal of alkanethiols by heterogeneous reaction with metal oxides is
presented. This method reduces the level of one category of sulfur compound to <20 ppm
by weight sulfur by selectively reacting the thiols with certain metal oxides, forming
metal thiolates. The metal thiolates are insoluble in hydrocarbons and water at
sufficiently low temperatures, permitting their removal by filtration. The metal oxide can
also be recovered by decomposing the thiolate with an oxidizing acid.
The second technique is the removal of sulfides and thiophene by oxidation using
sulfuric acid. Sulfuric acid is shown to be a fast and effective oxidizer of aromatic and
aliphatic sulfides, yielding sulfoxides and sulfones, which can then be extracted into a
polar phase. Both of these techniques, plus the study of polymer membrane reaction
mechanisms, are described in detail in this volume.
Chapter 2 Background
14
2. Background
2.1. Petroleum Refining
Refining crude oil is a complicated series of chemical processes designed to separate
petroleum into a variety of useful products, each meeting certain compositional criteria.
Refining begins by fractionating (distilling) crude oil into a series of streams with defined
boiling ranges. Table 2.1 shows some of the fractions and their boiling ranges.
Fuels, including gasoline, diesel, and kerosene (jet fuel), are the most valuable
products from petroleum. To enhance the quantity of these fuels produced from a single
barrel of crude, heavier streams are cracked, or broken down into smaller molecules.
Fluid catalytic cracking (FCC) typically utilizes a solid acid zeolite catalyst, often
promoted with rare earth metals5 in a fluidized bed. Large molecules are broken down to
create additional material in the naphtha range in order to produce more gasoline, a
valuable product. The “cracked naphtha” stream often contains larger amounts of sulfur
than virgin naphtha, since much of the sulfur in crude is in the form of heavy polynuclear
aromatic molecules present in the FCC feed stream.
Two additional processes are used to improve the quality of the resulting fuels,
particularly gasoline. Reforming uses Pt based catalysts to isomerize linear paraffins,
such as n-hexane, to higher octane number branched paraffins like 2,3-dimethylbutane. Pt
supported on chlorided alumina, sulfated zirconia, and zeolites are all used.6 The support
alters the activity of the catalyst, with alumina being most active and zeolites being least
Chapter 2 Background
15
active. However, high activity catalysts are more susceptible to poisoning by sulfur and
water.6 Removal of sulfur compounds before reforming gasoline streams is therefore
required.
The second process used to improve the quality of gasoline is alkylation. Alkylation
reacts n-butene with isobutane to create 2,2,4-trimethylpentane, also called isooctane, and
other branched paraffins.7 Alkylation also uses an acid catalyst, but due to excessive
coking, only liquid acid catalysts are currently used. Alkylation reactors blend either
sulfuric or hydrofluoric acid with the butane/isobutene stream to create alkylate, a high
quality gasoline that is blended into other gasoline streams.
The last major process used in oil refining is hydrotreating, or hydrodesulfurization
(HDS). Crude petroleum typically contains from 0.1 wt% to 3.0 wt% sulfur, depending
on the source. Crude from Norway and the North Sea typically have the lowest sulfur
concentration, while that from the Arabian Peninsula contains the most.8 Table 2.2 shows
the distribution of some aromatic sulfur species found in petroleum by boiling point. The
most common light species, denoted “gasoline-range sulfur” in the table, are methane-,
ethane-, and t-butanethiol, dimethyl sulfide, carbonyl sulfide (COS), and
tetrahydrothiophene.9
Chapter 2 Background
16
Table 2.1: Petroleum distillate fractions and their boiling points. Adapted from Pafko.10
Distillate Fraction Boiling Point (ºC) Carbon Number
Figure 3.8: Rate of hydrogenation of 1-decene over Pt and Pd catalysts (from Figure 3.4) compared
to Langmuir-Hinshelwood and Eley-Rideal style mechanisms. The fit of the power law ki0.5 is also
shown.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
50
The acetone hydrogenation results, shown in Figure 3.6, are not fit particularly well by
either of these rate laws. Instead, a better fit was obtained with a power law with
exponent 0.75 for both Pt and Cu. The exponent in this fit is perhaps a result of the
averaging of two competing mechanisms or multiple rate limiting steps, one with the 0.5
power seen with the olefin, and the other with a linear dependence on current. Such a
situation might arise from non-uniform distribution of protons arriving at the catalyst,
leading to some areas where hydrogenation is rate limiting (0.5 power) and others where
hydrogen supply is limiting (1.0 power).
It can be concluded from these results that the rate of hydrogen formation on the
surface of the catalyst is not the rate limiting step for current densities above 3 mA cm-2,
as suggested previously.58 If proton reduction were rate limiting, the plots in Figures 2-4
would be linear with current. All the data show that the rate increases less than linearly
with current, especially at high current density. This conclusion is further supported by
the absence of reaction on the ineffective catalyst metals. If hydrogen adsorption were
rate limiting, reaction would occur on all of the metals when current is applied.
The current density (or pressure for gas phase reactions) necessary to shift the reaction
from mass transport limited to reaction limited depends on the relative rates of mass
transport and surface reaction. A fast surface reaction will require large currents or
pressures to become reaction limited. In this study, the rates of hydrogenation on Pt and
Pd are accelerated by increasing the current density, suggesting that the reaction rate is of
a similar order of magnitude as the rate of mass transfer. On the ineffective catalysts,
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
51
however, the rate of reaction is independent of current, suggesting a much slower
reaction rate. When the rate of surface reaction is slow the adsorbed hydrogen primarily
reacts by recombination and desorption as H2. In the rate laws listed in Figure 5 this
corresponds to the first term on the right hand side of the rate laws being dominant. The
rate of mass transport is controlled by the current and is therefore the same on both the
active and inactive catalysts, but the reaction selectivity depends on the catalytic activity.
In a PEMHR the rate determining step and reaction selectivity changes with the current.
In catalytic reactions using gas phase hydrogen, the rate limiting step changes with
pressure; the overall reaction is limited by mass transfer of hydrogen to the surface
(adsorption) at low pressure and then shifts to surface reaction limited at higher pressures.
Increasing agitation speed in a slurry reactor can have the same effect of reducing mass
transfer limitations, causing a shift in rate limiting step.66,67 Pressure or mixing is
employed to alter the reaction rate in conventional hydrogenation reactors.
For example, the Langmuir-Hinshelwood rate law for a generic surface reaction A + B
→ C is
( )21
A A b BA
A A b B C C
b P b Pdnk
dt b P b P b P= −
+ + + (3.2)
where Px is the partial pressure of species x and bx are constants. This type of rate
equation is representative of hydrogenation reactions where A is the olefin, B is hydrogen
and C is the paraffin product. At low pressures of B, the rate appears first order in B. The
rate then passes through a maximum at moderate pressures of B. At higher pressures of
B, species B displaces A on the surface and the reaction rate decreases.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
52
Analogously, the rate-limiting step in PEM hydrogenation reactors may change from
being linear in current at low current density (relative to the rate of the surface reaction)
to less than linear at higher current density. At very high current density, the hydrogen
gas generation may displace the organic at the electrode surface, causing a decrease in
reaction rate. A maximum in the rate of hydrogenation was not observed here because the
rate of H2 generation became disruptive to the measurements before that point could be
reached.
Hydrogen surface coverage is the true controlling variable for catalytic hydrogenation.
In gas phase reactions the hydrogen pressure is manipulated to control hydrogen
coverage. In the PEMHR the current is manipulated to control the hydrogen coverage.
Ideally the rate should be related to the current supplied to the active catalyst surface.
Some of the proton current can also be reduced on the carbon electrode support, where
hydrogen diffusion and/or recombination can occur. Ideally the current density should be
based on the active catalyst area, as reaction rates are normalized to the active catalyst
area for supported metals. In a PEMHR, active catalyst must be accessible to both the
organic and the proton current to ensure that the two reactants mix on the surface.
Catalytic electrodes with extremely high surface areas, such as those consisting of
supported metal particles, may have a lower hydrogen surface coverage at a given current
than one of the same size with a smooth surface. High electrode surface area-to-current
ratios are equivalent to very low hydrogen partial pressures, since the true current density
(defined per unit surface area of the metal) is very low. Because of the complex structure
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
53
of the cathode/polymer electrolyte/hydrocarbon interface, good analytic tools to identify
the active catalyst surface are not available. We have followed tradition and reported
current density based the geometric area of the electrode.
3.1.5. Conclusions
Hydrogenation with a PEM hydrogenation reactor is similar to traditional catalytic
hydrogenation, with the added benefit of generating high surface coverages of hydrogen
at ambient pressure. Current density in a PEMHR is analogous to hydrogen partial
pressure in a conventional catalytic hydrogenation reactor. The rate of hydrogenation of
1-decene in a PEMHR is limited by the addition of one atom of hydrogen to the adsorbed
olefin rather than the rate of hydrogen adsorption onto the catalyst for all of the Pt and Au
group metals. The rate is well fit by a rate law in the form ki0.5, analogous to rates found
in conventional hydrogenation reactors. Aliphatic ketones hydrogenate much more
slowly than olefins at the experimental conditions. Acetone hydrogenated one order of
magnitude slower than 1-decene on Pt, and 4-heptanone did not react, suggesting a steric
effect blocking access to the central carbonyl. Acetone hydrogenation over Pt and Cu was
fit by ki0.75, suggesting a more complex mechanism for this reaction. Since the
hydrogenation mechanisms in PEMHRs are catalytic, the true hydrogen surface coverage
controls the rate of reaction rather than current density, which is often not defined in
terms of the surface area of catalyst but rather in terms of the geometric area of the
electrode.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
54
3.1.6. Acknowledgements
The author would like to thank Professor Weiyong Ying of the Department of
Chemical Engineering, East China University of Science and Technology, Shanghai,
China, who contributed to the experimental work described in this section.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
55
3.2. Electrochemical Reactions in a Polymer Membrane Reactor
3.2.1. Introduction and Theory
In the preceding sections, hydrogenation reactions occurring in a PEMHR were shown
to be catalytic. These reactions are exothermic and are thermodynamically favorable, but
rely on the interaction between the reacting molecules and the catalytic surface. Another
type of reaction is also possible in PEMHRs: electrochemical reactions. Electrochemical
reactions are driven by the potential at one of the two electrodes and can be
thermodynamically unfavorable (positive ∆Grxn).
The dehydrogenation of isopropanol is used to demonstrate the differences between
catalytic and electrochemical reactions. This reaction is the reverse of the hydrogenation
of acetone reaction discussed earlier. Like most dehydrogenation reactions, it has a
positive ∆Grxn = 27.6 kJ/mol and is driven by the positive potential at the anode. In the
current study, water electrolysis competes with isopropanol dehydrogenation. The charge
is balanced for both reactions by the reduction of protons at the cathode to form H2. The
mechanism for these reactions is given as Figure 3.9.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
56
1
3 71
2
3 7 2
3
3
Isopropanol M C H O M H e
C H O M Acetone M H e
Acetone M Acetone M
+ −
−
+ −
−
−
→+ + +←
→ + +←
→ +←
L
L L
L
4
2 4
5
5
6 1226
H O M HO M H e
HO M O M H e
O M O M
+ −
−
+ −
−
−
•
•
→+ + +←
→ + +←
→ +←
L
L L
L
Dehydrogenation of isopropanol
(oxidative adsorption of isopropanol)
(oxidation of adsorbed isopropoxide)
(desorption of acetone)
(oxidative adsorption of water)
(oxidation of adsorbed hydroxide)
(oxygen recombination and desorption)
Electrolysis of water
1
3 71
2
3 7 2
3
3
Isopropanol M C H O M H e
C H O M Acetone M H e
Acetone M Acetone M
+ −
−
+ −
−
−
→+ + +←
→ + +←
→ +←
L
L L
L
4
2 4
5
5
6 1226
H O M HO M H e
HO M O M H e
O M O M
+ −
−
+ −
−
−
•
•
→+ + +←
→ + +←
→ +←
L
L L
L
Dehydrogenation of isopropanol
(oxidative adsorption of isopropanol)
(oxidation of adsorbed isopropoxide)
(desorption of acetone)
(oxidative adsorption of water)
(oxidation of adsorbed hydroxide)
(oxygen recombination and desorption)
Electrolysis of water
Figure 3.9: Reaction mechanism for the dehydrogenation of isopropanol and electrolysis of water.
The removal of the first electron is rate limiting in each mechanism.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
57
Assuming the removal of the first electron and proton from isopropanol and water is
the rate limiting step, the mechanisms shown in Figure 3.9 suggest the rate law shown in
equation (3.6). The expression derived states that the rate of dehydrogenation of
isopropanol is the total current times the selectivity for the isopropanol reaction over the
water reaction.
iso iso T M isor = k [Isopropanol]M exp( (i)/(RT)) θ α εF (3.3) W W T M Wr = k [Water]M exp( (i)/(RT)) θ α εF (3.4) iso iso iso waterr = (i/(2 ))(r /(r +r )) F (3.5)
( )2
2[ ] ( )
1 exp[ ]
isowater water iso
iso
i
rk H O i
RTk Isopropanolα α ε
=
− +
FF
(3.6)
where F is Faraday’s constant, ε is the potential of the anode, i is the current density, and
α is the asymmetry factor which accounts for differences between the forward and
reverse reaction mechanisms.
Equation (3.6) suggests that the rate of dehydrogenation should be a linear function of
current if the two asymmetry factors are the same. The asymmetry factor for any reaction
is 0.5 if the reaction has the same forward and reverse reaction mechanism. Also, if the
reaction is electrochemical, the anode material should not affect the reaction rate, in
strong contrast to a catalytic reaction.
3.2.2. Experimental
The dehydrogenation reaction was performed in the PEMHR using 10 vol%
isopropanol in DI water, with 0.1 M H2SO4 as the supporting electrolyte. The same
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
58
solution was used on both sides of the reactor due to the high permeability of isopropanol
through Nafion. The reaction was performed over Pt mesh and carbon cloth anodes using
a Pt mesh cathode. The reaction temperature was 50°C and reaction time was 1 hour. The
same experimental setup and protocols were used as for the acetone hydrogenation
described previously. Product analysis was performed by gas chromatography.
3.2.3. Results and Discussion
Figure 3.10 confirms the electrochemical nature of this reaction. On the Pt anode, the
rate of dehydrogenation of isopropanol is linear with current, as predicted by equation
(3.6). On the C anode, the rate is linear at low current but reaches a maximum at higher
current. The reaction rate over C is also predicted by equation (3.6) by assuming that αiso
? αw. The difference in the asymmetry factors over C likely arises from a surface
interaction of the isopropanol or an intermediate with the carbon causing the mechanism
of the dehydrogenation to change slightly. Figure 3.10 also demonstrates how the
electrode material has little effect on the rate of the reaction. Until the asymmetry
difference becomes dominant at high current, the rate of dehydrogenation over Pt and C
are equal.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
59
0
10
20
30
40
50
60
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
Current Density (mA/cm2)
Rea
ctio
n R
ate
(nm
ol/c
m2 /s
ec)
0.0
2.0
4.0
6.0
8.0
10.0
Rea
ctio
n R
ate
(mA
/cm
2 )
Experimental - Pt
Experimental - C
Predicted - C
Predicted - Pt
Figure 3.10: Rate of dehydrogenation of isopropanol over Pt and C anodes. Lines are fits to EQ.
Fitting parameters for Pt are kw/kiso = 0.0135, αw-α iso = 0. Parameters for C are kw/kiso = 0.0102, αw-
α iso = 0.026.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
60
The dehydrogenation of isopropanol clearly demonstrates the difference between a
catalytic and an electrochemical reaction. Electrochemical reactions are less sensitive to
electrode material and surface area and the rate limiting step involves a charge transfer.
Because charge transfer is rate limiting, more applied current yields a higher reaction
rate, unless asymmetries exist between competing reactions, as observed with the carbon
anode.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
61
3.3. Desulfurization in a Polymer Membrane Hydrogenation Reactor
Numerous attempts were made to achieve desulfurization reactions in a polymer
membrane reactor. The most reactive species, alkanethiol, was tested in a PEMHR using
Pt, Pd, Cu, Al, Pb, Fe, C, Co-Mo (deposited on C from the salts), and Ag electrodes.
Solutions of 10 vol% dodecanethiol in cyclohexane were reacted at current densities from
1-15 mA cm-2, and also under cyclic voltage and current programs.
Adsorption of thiol was evident on high surface area carbon electrodes, and a trace
amount of the desulfurization product, dodecane, was detected. The production of
dodecane from activated carbon electrodes probably results from surface adsorption and
reaction. No application of current or voltage was able to regenerate the surface to allow
the reaction to continue once the initial trace amounts of thiol had adsorbed and reacted.
This result is consistent with the difficulty of regenerating high surface area adsorbents
for desulfurization, which typically require calcination.68
Pb electrodes, which are capable of high energy electroreductions69, also were unable
to reduce thiols. The only product observed was lead thiolate, described in the following
chapter. Cu electrodes produced a small amount of a waxy polymeric material, but no
dodecane was produced. The other metals were completely ineffective in reducing thiols
under any conditions.
Cyclic voltammetry was performed for Cu, Ni, Pb, Ag, Pd, and Pt electrodes to
determine if any reactions involving an electron transfer, such as a direct electroreduction
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
62
of thiol, were occurring. Electrodes of each metal were submersed in solutions of
approximately 60 vol% ethanol and 36 vol% deionized water. The balance of the solution
is sulfuric acid (supporting electrolyte) and the sulfur compound of interest. The counter
electrode was a Pt wire. Potential was controlled with an Arbin MSTAT4+ battery testing
station. The potential was referenced against a Pt wire in the same electrolyte solution.
The potential window was experimentally determined for each compound by scanning
until the current exceeded about 10 mA.
No irreversible reduction of thiol was observed for any of the studied electrodes,
although Ag would react with some sulfur compounds to form a non-adherent layer of
Ag2S. Unfortunately, the metal was not able to be completely regenerated by
electrochemical means, leading to the disintegration of the electrode. Figure 3.11 shows
the cyclic voltammogram of an Ag electrode in an electrolyte solution containing
tetrahydrothiophene. The difference in size of the anodic (downward) and cathodic
(upwards) peaks is the amount of metal that cannot be regenerated. The use of a Ag
electrode in the PEMHR was unsuccessful in producing any desulfurization products.
While sulfur compounds were observed to interact with several of the metals in a
potential-dependent manner, the lack of a reduction reaction occurring at the surface,
either catalytic or electrochemical, suggests that PEMHRs are not suited to
desulfurization reactions. The high temperatures and specific catalysts required for HDS
are not possible in a PEMHR. Additionally, economics requires that any desulfurization
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
63
reaction occur quickly to minimize the necessary reactor area. Other techniques for
desulfurization are therefore required.
Chapter 3. Reaction Mechanisms in Polymer Electrolyte Membrane Hydrogenation Reactors
64
-10
-8
-6
-4
-2
0
2
4
-1500 -1000 -500 0 500 1000 1500
Voltage (mV)
Cur
rent
(mA
)
Low S ConcMed S ConcHigh S Conc
Anodic Current
Figure 3.11: Cyclic voltammogram of an Ag electrode in a tetrahydrothiophene solution. The anodic
(oxidation) current increases with sulfur concentration, indicating the production of more Ag2S. The
smaller cathodic wave near 0 V is the reduction of some of the Ag2S.
Anodic current increases with increasing sulfur compound concentration
Reduction of Ag2S increases slowly with increasing sulfur concentration
Chapter 4 Removal and Recovery of Alkanethiols
65
4. Removal and Recovery of Alkanethiols
Thiols are present in petroleum from many parts of the world. The “sweet” designation
of crude refers to low thiol content, while “sour” petroleum has high concentrations of
thiols. Alkanethiols are also produced in HDS reactors by the reaction of olefins with H2S
present in the reactor. Here we present a new process for the removal and recovery of
alkanethiols from petroleum.
4.1. Chemistry and Reactions
4.1.1. Introduction
The removal of thiols from hydrocarbon streams, such as petroleum, has been a
concern since the earliest days of refining. Thiols have a strong and unpleasant odor and
also can be acidic. The removal of thiols from petroleum is commonly called sweetening.
Most crude oils contain some thiol species, and a few contain significant quantities of
thiols.70
Over the past fifty years, the removal of sulfur from petroleum products has become an
essential part of the refining process. Sulfur poisoning of Pt reforming catalysts led to the
widespread use of hydrodesulfurization (HDS).14,71 HDS reacts organosulfur compounds
with hydrogen over a catalyst at elevated temperature (T = 300-450 ºC) and pressure (P =
35-270 bar) to produce a hydrocarbon and H2S.14 The feedstocks treated by HDS are
Chapter 4 Removal and Recovery of Alkanethiols
66
primarily those used for catalytic reforming to make gasoline. Desulfurization of diesel
fuel is now required to meet new environmental emissions standards.72
The reaction described here provides the basis for a simple, low temperature method
for the removal of thiols from a hydrocarbon stream. In this reaction, thiols are converted
directly to insoluble metal thiolates by heterogeneous reaction with certain metal oxides
or hydroxides. The thiolates can then be mechanically filtered from the hydrocarbon
stream. The metal and the original thiols can then be recovered in a reactive liquid-liquid
extraction step. This reaction could form a process to complement conventional HDS by
reducing the sulfur load on an HDS reactor while recovering the thiols for other uses.73-75
The reaction is based on thiolate chemistry in which thiols react with metal ions to
form metal thiolates.76,77 Most work in this area focuses on the reaction of thiols with
aqueous solutions of heavy metal salts77-79 or with surface reactions on metals80-84 and
metal oxides.85-87 The reaction system applied here is distinct from these works because
the reaction removes ions from the surface of the oxide, allowing a dense oxide particle
to be completely consumed. While the heterogeneous reaction was almost certainly
observed previously, the reaction is absent from much of the literature.88,89 Sources
indicating that such a reaction occurs do not specify the reaction.76 We demonstrate here
some preliminary results of the heterogeneous reaction between a liquid thiol and a solid
metal oxide to form a solid thiolate and water, and apply this obscure reaction to the
modern day problem of removing and recovering thiols from a hydrocarbon mixture. An
example reaction with PbO is
Chapter 4 Removal and Recovery of Alkanethiols
67
2 2PbO + 2 RSH Pb(SR) + H O→ (4.1)
The reaction works with any combination of thiols and metal oxides that form stable,
insoluble thiolates. We have found numerous oxides and hydroxides of Pb, Hg(II), and
Ba to react with alkanethiols.
The reaction of thiols with these metal oxides or hydroxides occurs quickly and
completely at room temperature and atmospheric pressure. Under these conditions, the
thiolates are solid and insoluble in either aqueous or hydrocarbon liquids, consistent with
older studies of Pb thiolates.78
After separation of the thiolates from the hydrocarbon solution, the recovery of the
metals is achieved by reactive extraction with a dilute oxidizing acid. The choice of acid
depends on the metal salt desired. An example reaction is
2 3 3 2Pb(SR) + 2 HNO Pb(NO ) + 2 RSH→ (4.2)
This simple process uses inexpensive materials and very little energy compared to
conventional desulfurization techniques.
Potentially reactive oxides are identified through simple thermodynamic
considerations. The reaction of a thiol with a metal oxide involves the breaking of an S-H
and an M-O bond and the formation of an M-S and an O-H bond per reacting thiol
molecule. An estimate of ∆Hrxn can be obtained by comparing the heats of formation of
Chapter 4 Removal and Recovery of Alkanethiols
68
the metal oxide and the metal sulfide. This technique identifies which reactions are
favorable but does not quantitatively predict the kinetics of the reaction.
4.1.2. Experimental
Powdered PbO (massicot form, Aldrich) of mass between 0.02 and 0.2 g was added to
a glass vial. An n-alkanethiol, either 1-hexanethiol, 1-octanethiol, or 1-dodecanethiol
(Aldrich), was added to the vial. An inert hydrocarbon, usually cyclohexane or toluene,
was sometimes added to prevent the reaction product from becoming a paste. The solid
product could be suspended in the hydrocarbon by stirring to facilitate filtration. Some
samples were stirred using a magnetic stir bar, and some were heated to about 70°C on a
hotplate and cooled before filtering.
The solid reaction product, a yellow powder, was gravity filtered (Whatman #50
hardened circles), then rinsed several times with water, acetone, cyclohexane, and then
pentane and dried at room temperature.
The Pb was extracted from the thiolates by liquid-liquid extraction with dilute (0.2-0.3
M) nitric acid. The solid thiolates were added to cyclohexane and the mixture was heated
to the melting point of the thiolate, approximately 60-80 ºC. This mixture was added to
the nitric acid and agitated by rapid stirring. The hydrocarbon layer was then analyzed for
thiol content by gas chromatography (Hewlett Packard HP-1 column). The aqueous layer
was evaporated and the Pb was recovered as Pb(NO3)2 crystals.
Chapter 4 Removal and Recovery of Alkanethiols
69
Other metal oxides and hydroxides were tested by adding 1-octanethiol to a small
amount (approx. 0.15-0.5 g) of the oxide or hydroxide in a glass vial. The formation of
thiolates was evidenced by a dramatic increase in apparent solid volume and in some
cases, a color change. Reacted mixtures were filtered and dried as above, then weighed to
determine conversion. The compounds tested were BaO, Bi2O3, CaO, CdO, CoO, Cr2O3,
sulfur (~0.047 M) as tetrahydrothiophene in tetrahydrofuran (Alfa Aesar, 99+%), (3)
2050 ppmw sulfur (0.056 M) as thiophene (Aldrich, 99+%) in toluene (Sigma-Aldrich,
99.5%) and (4) dibutyl sulfide (Aldrich, 96%) in toluene. A small amount of decane
(Aldrich, 99+%) was added to each stock solution as an internal standard for GC analysis.
Ten mL of the organic solution and 10 mL of sulfuric acid (dilutions of 96% H2SO4,
Fisher) were added to the reaction vessel, forming two liquid phases. The reaction
temperature was 22ºC. The two phase mixture was vigorously stirred for 1 hour (unless
otherwise indicated). After reaction, the phases were allowed to separate and the organic
layer was removed by syringe to to prevent further reaction.
The concentration of the sulfide species in the organic layer was measured by gas
chromatography (30m methyl silicone capillary column, flame ionization detector). The
concentration of unreacted sulfide species was determined by normalizing all results
using the internal standard (decane). We were unable to accurately analyze the
composition of the aqueous layer by GC due to the high acid content. Compositions of
organic species in the aqueous phases are only semiquantitative. To analyze the aqueous
phase by GC the solutions were diluted to reduce the acid concentration. However, the
acid either adsorbed or reacted with the GC column packing and results were only
semiquantitative. Quantitative analysis was limited to the organic phase.
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
116
Tetrahydrothiophene also was reacted in a homogeneous solution of tetrahydrofuran
and sulfuric acid. The reaction products obtained by this technique were analyzed by
mass spectrometry.
5.4. Results
All three sulfide compounds were successfully removed from the hydrocarbon solvent
by oxidation and extraction into the aqueous (acid) phase. No oxidized reaction products
remained in the organic phase, and only trace amounts (<50 ppm) of the sulfide
compounds were detected in the aqueous phase. Colloidal elemental sulfur was formed in
the aqueous phase, evidenced by the classic Tyndall “sunset” effect, shown in Figure 5.2.
The aqueous phase became yellow upon mixing with the organic phase. As the reaction
proceeds, the aqueous phase color changed first to orange, then brownish-red. If high
concentrations of sulfide were used, the colloid concentration became large enough to
eventually turn dark red and finally black, as no light was transmitted. The aqueous phase
containing the colloidal sulfur was recovered by gravity separation. This colloidal
solution became milky white if the acid was diluted sufficiently with water, forming
“milk of sulfur”. The colloidal sulfur could be slowly dissolved with CS2, yielding an
optically clear aqueous phase.
Gas chromatography analysis of the aqueous phases revealed three primary oxidized
sulfide products for the tetrahydrothiophene reaction. All three oxidized products were
extracted completely into the aqueous phase, there was no evidence for any of these
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
117
products in the organic phase. The same three products were detected by GC for the
reaction of tetrahydrothiophene in THF as in the diluted aqueous phase. The products
were analyzed by mass spectrometry, shown in Figure 5.3. The mass fragmentation
patterns were compared to those in the NIST database (http://webbook.nist.gov). The
spectrum of product B matches tetrahydrothiophene sulfoxide. Product A was not able to
be positively identified, but based on the parent peak and fragmentation the compound is
likely a ketosulfoxide species or the dehydrogenated sulfone. Product C yielded a highly
fragmented spectrum with mass fragments going up to m/z=189. The mass spectrum
appears best described as an adduct of tetrahydrothiophene and sulfuric acid, shown in
figure 2.
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
118
Figure 5.2: Photographs of the Tyndall sunset effect in a solution of toluene, tetrahydrothiophene,
and thiophene.
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
119
The reaction rate was measured for each of the three sulfides when reacting with 16M
H2SO4 (18 M sulfuric acid is fully concentrated (96 wt% H2SO4)). The stirring was
stopped for several seconds after various time intervals and the phases allowed to
separate. Samples were drawn from the organic phase and the concentration of unreacted
sulfide was determined by GC. The results are shown in Figure 5.4; the observed reaction
rates are well fit by a first order rate law. Dibutyl sulfide was removed to about 27 ppm S
within 5 minutes and was undetectable (<10 ppm S) at 20 minutes. Tetrahydrothiophene
was reduced to 120 ppm in 5 minutes, and reached 66 ppm S after 65 minutes. Thiophene
reacted more slowly, being reduced to 283 ppm S at 20 minutes and 113 ppm S after 65
minutes. Values below about 150 ppm S are near the limit of reliable quantification for
the analysis method used.
The reaction rate appears to be first order with respect to sulfuric acid concentration at
lower acid concentrations. Figure 5.5 shows the integrated reaction rate of the
tetrahydrothiophene solution over 1 hour for various acid concentrations. Figure 5.6
shows the extent of reaction over time for the same system using 9M acid. The reaction
rate shown in Figure 5.6 is constant over time, suggesting that the reaction is now zero
order in sulfide concentration. Considering Figure 5.5 and Figure 5.6, the lower acid
concentration appears to shift the reaction from being first order in sulfide concentration
to first order in sulfuric acid concentration. A reaction scheme consistent with these
results is presented in the following section.
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
120
A
B
C
S
O
-200 0 200 400 600 800 1000 1200 1400 16000.0
0.1
0.2
0.3
0.4
0.5
Sig
nal (
V)
Time (sec)
Solvent
THT
A
B
C
S
O
OS
OO
or
S
OS
O
O
O (H+)
A
B
C
S
O
-200 0 200 400 600 800 1000 1200 1400 16000.0
0.1
0.2
0.3
0.4
0.5
Sig
nal (
V)
Time (sec)
Solvent
THT
A
B
C
S
O
OS
OO
or
S
OS
O
O
O (H+)
Figure 5.3: GC/MS spectra for the reaction products of tetrahydrothiophene and sulfuric acid. The
mass spectrum for product B matches that of tetrohydrothiophene 1-oxide. The spectrum of product
A suggests a species with two oxygen atoms, such as the sulfone or ketosulfoxide. Mass spectrum C
could be the adduct of sulfuric acid and tetrahydrothiophene.
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
121
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60 70
Reaction Time (min)
Rem
aini
ng S
ulfu
r (p
pm S
)
? Dibutyl Sulfide? Thiophene× Tetrahydrothiophene
Figure 5.4: Reaction rate of 2000 ppm solutions of thiophene, tetrahydrothiophene, and dibutyl
sulfide with 16M H2SO4. Data are fit by a rate law that is first order in sulfide concentration, shown
above. For the given rate constants, sulfide concentration is given in units of ppm S.
[ ] [ ]
1
1
1
0.098min
0.559min
0.868min
Thiophene
Tetrahydrothiophene
Dibutylsulfide
d Sulfiderate k Sulfide
dt
k
k
k
−
−
−
−= =
=
=
=
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
122
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 2 4 6 8 10 12 14 16 18
H2SO4 Concentration (M)
% T
HT
Rem
ove
d (
% o
f in
itia
l)
Figure 5.5: Consumption of tetrahydrothiophene from a 2000 ppm S solution after 1 hour reaction
with various concentrations of sulfuric acid. Data is fit by a rate law first order in acid concentration
shown. [H2SO4*] is acid concentration above the threshold value shown by the intercept (2.4 M).
[ ] [ ]2 4
1
*
0.0712minTHT
d THTrate k H SO
dt
k −
−= =
=
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
123
0%
10%
20%
30%
40%
50%
60%
0 10 20 30 40 50 60 70 80 90 100
Reaction Time (min)
TH
T R
emov
ed (%
of i
nitia
l)
Figure 5.6: Reaction rate of tetrahydrothiophene from a 2000 ppm S solution reacting with 9M
H2SO4. Rate is constant at 0.0055 min-1.
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
124
5.5. Discussion
A reaction scheme developed for the reaction of H2S with sulfuric acid can be easily
adapted to explain the observed results. Chuang, et al118,119 have shown that a two step
reaction, consisting first of reaction between H2S and H2SO4, produces SO2, which is
then consumed by further reduction to elemental sulfur. If a similar scheme is assumed
for sulfides, the resulting reactions are:
(5.1) (5.2)
where the dimethyl sulfide compounds shown can be replaced by any sulfide.
Using dimethyl sulfide as an example in the above reactions, the Gibbs energy of
reaction (5.1) is +25.5 kJ mol-1, using each material in its standard state.120 However, in
concentrated acid solutions, the heat of mixing of water and sulfuric acid is considerable.
The partial molar enthalpy of formation of water in a 96 wt% solution of sulfuric acid is
higher by about 30 kJ/mol than the standard state121, making the overall ∆Grxn ˜ -5
kJ/mol. Other sulfide compounds have approximately the same free energy change for
the formation of the sulfoxide from the sulfide. The large excess of sulfuric acid and the
consumption of SO2 by reaction (5.2) favor formation of the product sulfoxide. No gas
was ever observed being evolved from the reaction, probably due to the very high
solubility of SO2 in the concentrated acid. The solubility of SO2 in sulfuric acid ranges
∆Grxn ˜ -5 kJ mol-1
∆Grxn = -45.5 kJ mol-1
S S
O
S SOOO
+ H2SO4+ H2O + SO2
+ SO22 2 + S
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
125
from 10-100+ g SO2/kg H2SO4, depending on concentration and temperature. At the
experimental conditions, the solubility is about 30 g SO2/kg H2SO4. SO2 is also soluble in
most organic liquids.122
Several other reactions are also possible, though most are not thermodynamically
favorable. One reaction that might occur with tetrahydrothiophene is a modification of
reaction (5.1), shown below. This reaction explains the presence of product A from
Figure 5.3. Note that the double bond can be formed at either of the two possible
positions with about the same energy. The formation of the ketosulfoxide would occur
from electrophilic attack on the ring, with the second oxygen adding at the alpha carbon,
rather than the sulfur.
(5.3)
These proposed reactions are thermodynamically favorable, account for the observed
products and also agree with the observed reaction rate laws. Figure 5.4 shows that at
high acid concentrations, the rate of oxidation is first order in sulfide concentration, in
agreement with reactions (5.1) and (5.3). The initial reaction between the sulfide and
H2SO4 appears to be rate limiting. This result agrees with the rate law determined for H2S
reacting with sulfuric acid.119
The active acid species has been proposed to be H2SO4, rather than HSO4-.119 At high
acid concentrations, the H2SO4 species is abundant and reaction 1 is the rate limiting step.
S SO O
+ H2SO4+ 2 H2O + S ∆Hrxn = -85.1 kJ mol-1
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
126
At lower acid concentrations, however, the dissociated species is more prevalent. The
reaction rate may then become limited by the rate of association of H2SO4 from the ions,
H+ + HSO4- → H2SO4, which must occur before the oxidation step can proceed. Under
these conditions, the rate would appear first order in acid concentration and zero order in
sulfide concentration, despite the acid being far in excess of the sulfide. The data in
Figure 5.6 supports this hypothesis, showing that the reaction rate in 9M H2SO4 is
constant (since the concentration of H2SO4 is not appreciably changed by the reaction
with small amounts of sulfide), and Figure 5.5 confirms the reaction is first order in acid
concentration. Below the threshold acid concentration shown in Figure 5.5, the rate of
association is likely too small to permit the oxidation reaction and the thermodynamics of
the reaction become unfavorable.
The reaction of sulfide species with concentrated sulfuric acid was previously reported
by Meadow, et al123,124, who attributed the desulfurization effect to NO2 additives in
small quantities of H2SO4. However, their analysis method could not distinguish between
sulfoxides and other oxidized sulfur species present in the hydrocarbon phase and
unreacted sulfide, nor did they report the reduction of sulfuric acid.
Early work of Wood et al suggested that sulfides were removed by dissolution into
sulfuric acid.125 In the present work, all of the oxidized sulfur compounds were extracted
into the aqueous phase, while the unreacted sulfides remained in the organic phase. The
octanol-water partition coefficients126 for the three sulfides studied here are summarized
in Table 5.1. All three sulfides show a strong preference for the organic phase.
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
127
Furthermore, very little of the original sulfide was detected in the acid phase after the
reaction. The absence of the sulfide species and the presence of the sulfoxides and other
oxidized species reveal that the observed “solubility” only occurs after the sulfides are
converted to the more polar oxidized species.
Table 5.1: Partition coefficients for the sulfides studied. Data fromYaws.126
Species Log of Octanol-Water
Partition Coefficient (25°C)
Thiophene 1.81
Tetrahydrothiophene 1.79
Dibutyl sulfide 3.87
This simple extraction method reported here may not work with higher molecular
weight sulfides. The higher molecular weight sulfoxides and sulfones may not be
sufficiently polar to be fully extracted into the aqueous phase. These higher molecular
weight species may require the sulfoxides and sulfones to be extracted with methanol or
adsorbed onto alumina to effect their removal, as is done in conventional peroxide-based
ODS.37
The ability of sulfuric acid to quickly remove sulfide species by converting them into
sulfoxides and other oxidized species provides a possible addition to or replacement for
ODS performed by hydrogen peroxide. Since peroxide-based ODS is limited by the cost
of the oxidant, a pre-treatment with sulfuric acid could remove much of the original
sulfur content before the stream is sent to the ODS reactor. A sulfuric acid wash could
also reduce the load on hydrodesulfurization (HDS) reactors when processing high-sulfur
Chapter 5 Oxidation of Aliphatic and Aromatic Sulfides Using Sulfuric Acid
128
crude. The reactor design for such a sulfuric acid system is essentially the same as for
sulfuric acid-based alkylation reactors. Only low olefin feeds could be used since the acid
will catalyze polymerization reactions.
5.6. Conclusions
Oxidative extraction of thiophene, tetrahydrothiophene, and dibutyl sulfide was
demonstrated in a two phase reactor using concentrated sulfuric acid. The oxidized
products are likely the sulfoxides and other oxidized sulfur species, which are effectively
extracted into the acid phase. Sulfuric acid is reduced to elemental sulfur, which remains
in a colloidal state in the aqueous phase. The reaction rates are first order in sulfide
concentration when performed in concentrated (16M+) sulfuric acid, and first order in
sulfuric acid when performed in less concentrated acids. This reaction could be used to
greatly reduce the sulfur concentration in petroleum fractions prior to HDS or ODS
processes.
Chapter 6 Conclusions
129
6. Conclusions
Despite extensive research into desulfurization by the petrochemical industry over the
last thirty years, few viable alternatives to hydrodesulfurization have been developed.
The lack of progress is likely due to a combination of factors including the effectiveness
of HDS, the cost associated with installing new processes, and resistance to change in the
established community of refineries.
Regulatory limitations on sulfur levels in automotive fuels, particularly diesel, have led
to a renewed interest in alternative desulfurization technologies, as demonstrated by the
commercial development of several oxidative desulfurization technologies. The scope of
new technologies under development is still limited, however. For example, virtually all
ODS research is based on hydrogen peroxide-organic acid systems.
The need exists for a wider variety of desulfurization techniques; particularly as
refineries are built in many parts of the world and a wider variety of crude sources are
used. The increasing variety of crude compositions, including the use of high-sulfur
“sour” crude from Russia and Canada, poses a particular challenge to refiners. The
existence of multiple desulfurization techniques would allow the efficient removal of
sulfur from these petroleum sources.
Despite the effectiveness of other catalytic processes such as HDS in sulfur removal,
PEMHRs were not successful. The reactions occurring in PEMHRs are catalytic and
follow the same rate laws as those occurring in conventional hydrogenation reactors.
Chapter 6 Conclusions
130
Except perhaps at very low current densities, the reaction step (addition of first hydrogen)
becomes rate limiting. Despite the similarities to HDS, the temperature, pressure, and
choice of catalysts available in PEMHRs are insufficient to effectively remove sulfur
from petroleum.
The removal of thiols from petroleum using metal oxides is very promising. PbO is
able to scavenge alkanethiols from a hydrocarbon solution to very low (~10 ppm S)
levels. The reaction is catalyzed by water, which likely functions as a proton transfer
agent during the reaction. The reaction is zero order in thiol concentration and
proportional to water concentration, permitting the fast removal of small amounts of
thiols. This process could complement HDS by performing a post-treating function to
remove recombinant thiols formed in the HDS reactor, minimizing the amount of metal
oxide used and thiolate formed. The product thiolates can also be easily decomposed to
regenerate the original thiols and metal oxide.
The primary limitations of this process are its specificity to thiols and the use of heavy
metals. The specificity of the reaction prevents this process from being applied
independently from an HDS unit unless the feed stream contains almost exclusively thiol
sulfur. The use of heavy metals, such as Pb, is also undesirable due to the risk of
contamination of the refinery site or the environment. The use of Bi instead of Pb can
mitigate this problem somewhat, although containment will remain an issue.
Chapter 6 Conclusions
131
The effectiveness of this process on actual crude or gasoline streams needs to be
investigated. These streams contain a wide variety of thiols and other organic compounds
that may complicate the removal of the thiols. Retail gasoline is not suited for this type of
analysis because it has already been hydrotreated to remove all of the thiols.
The use of sulfuric acid as a selective oxidizing agent for desulfurization of sulfides
and thiophene derivatives is also promising. This process relies on the thermodynamic
equilibrium between sulfur in the +6 oxidation state (sulfate) and sulfur in the -2
oxidation state (organic sulfide). Despite the stability of the sulfate ion, a mixture of these
two oxidation states will try to move towards equilibrium, generating the 0 (elemental)
and +4 (dioxide) oxidation states of sulfur. At sufficiently high concentrations of sulfuric
acid, sulfides and thiophene are oxidized to sulfoxides and other oxidized species in a
reaction that is first order in sulfide concentration until an equilibrium concentration is
reached.
The oxidation of the sulfides permits selective extraction or adsorption of the sulfur
compounds, identical to ODS. The effectiveness of this extraction is demonstrated by the
removal of the product species to undetectable levels in the hydrocarbon phase after
reactive extraction with concentrated sulfuric acid. This process replaces oxidation by the
expensive hydrogen peroxide with inexpensive sulfuric acid, without significant changes
to the remaining ODS flowsheet.
Chapter 6 Conclusions
132
The oxidation of sulfur compounds with sulfuric acid has two limitations. First, the
reaction reaches equilibrium for some compounds at sulfur levels that are above the legal
limits, requiring removal of the products and long reaction times. This limitation can be
overcome by combining the most desirable features of sulfuric acid-based ODS and
hydrogen peroxide-based ODS. A pretreatment with sulfuric acid can cheaply remove
most of the sulfur from petroleum streams. A peroxide unit can then oxidize the
remaining traces of sulfur to meet regulatory limits. The two units can share many of the
extraction, adsorption, and neutralization process steps.
The second limitation is the use of a liquid sulfuric acid. Extensive efforts are being
made to eliminate the use of sulfuric acid as a catalyst for alkylation due, in part, to the
need to dispose of large quantities of petroleum-contaminated acid. The introduction of
the same material into a new process is somewhat counterproductive, although the
technology for safely using and disposing of such acid already exists. Also, this process
could only be used on streams that do not contain significant amounts of olefins, since
sulfuric acid will catalyze alkylation and oligomerization reactions.
The effect of this treatment technique on actual diesel streams should be evaluated.
Specifically, the sulfur levels obtained and any side reactions occurring are of particular
importance. Methods to regenerate the sulfuric acid and remove the colloidal sulfur are
also needed before commercialization.
Chapter 6 Conclusions
133
Two new desulfurization techniques have been demonstrated here, each applicable to
different aspects of refining. Each permits the more efficient use of other desulfurization
technologies by the pre- or post-treatment of streams desulfurized by ODS or HDS.
Application of these technologies could permit the efficient production of ultra-low sulfur
fuels from lower quality, high-sulfur crude.
Chapter 7 References
134
7. References
1. Office of Energy Markets and End Use, World Energy Use and Carbon Dioxide Emissions, 1980-2001. 2004, US Energy Information Administration.
2. Clean Air Markets Division, EPA Acid Rain Program 2003 Progress Report. 2004, US EPA Office of Air and Radiation. p. 20 pages.
3. Office of Air Quality Planning & Standards, SO2 - How Sulfur Dioxide Affects the Way We Live & Breathe. 2000, US EPA.
4. Rao, T., Effects of Sulfur on Exhaust Emissions. 1997, US EPA - Office of Mobile Sources: Ann Arbor, MI.
5. Fowler, R. and L. Boock. AdVanta FCC Catalyst. in Grace 2002 Latin American and Caribbean Refining Seminar. 2002. Lima, Peru.
6. Beck, J. EMICT, A Regenerable Paraffin Isomerization Catalyst. in MCCJ 15th Anniversary Symposium. 2000. Japan.
7. Ackerman, S., G.K. Chitnis, and D.S. McCaffery, Jr., Advances in Sulfuric Acid Alkylation Process Improve Gasoline Blending Economics in World Refining, 7, 2002,
8. Irion, W.W. and O.S. Neuwirth, Oil Refining, in Ullmann's Encyclopedia of Industrial Chemistry. 2003, Wiley-VCH: Weinheim. p. 205-256.
9. Firor, R. and B. Quimby, Determine low-level sulfur in hydrocarbon gases in Hydrocarbon Processing, February, 2003, pp.79-81.
10. Pafko, W., Case Study: Petroleum, Modern Refining, 2000, http://www.pafko.com/history/h_refine.html.
11. Shiflett, W.K. and L.D. Krenzke, Consider improved catalyst technologies to remove sulfur in Hydrocarbon Processing, February, 2002, pp.41-43.
12. Knudsen, K.G., B.H. Cooper, and H. Topsøe, Catalyst and process technologies for ultra low sulfur diesel. Applied Catalysis A: General, 1999. 189: p. 205-215.
14. Gates, B.C., J.R. Katzer, and G.C.A. Schuit, Chemistry of Catalytic Processes. McGraw-Hill Chemical Engineering Series. 1979, New York: McGraw-Hill. 464.
Chapter 7 References
135
15. Sweed, N.H. SCANfining for Low Sulfur Gasoline. in Latin American and Caribbean Refining Seminar. 2002. Lima, Peru: ExxonMobil Research and Engineering Co.
16. Lauritsen, J.V., et al., Atomic-Scale Structure of Co-Mo-S Nanoclusters in Hydrotreating Catalysts. Journal of Catalysis, 2001. 197(1): p. 1-5.
17. Kishan, G., J.A.R. van Veen, and J.W. Niemantsverdriet, Realistic Surface Science Models of Hydrodesulfurization Catalysts on Planar Thin-Film Supports: The Role of Chelating Agents in the Preparation of CoW/SiO2 catalysts. Topics in Catalysis, 2004. 29(3 - 4): p. 103-110.
18. Kim, T.S., H.Y. Kim, and B.H. Kim, Petroleum Desulfurization by Desulfovibrio-Desulfuricans M6 Using Electrochemically Supplied Reducing Equivalent. Biotechnology Letters, 1990. 12(10): p. 757-760.
19. Borole, A.P., et al., Comparison of the emulsion characteristics of Rhodococcus erythropolis and Ecsherichia coli SOXC-5 cells expressing biodesulfurization genes. Biotechnology Progress, 2002. 18(1): p. 88-93.
20. Gray, K.A., G.T. Mrachko, and C.H. Squires, Biodesulfurization of fossil fuels. Current Opinion in Microbiology, 2003. 6(3): p. 229-235.
21. Philip, L. and M.A. Deshusses, Sulfur dioxide treatment from flue gases using a biotrickling filter - bioreactor system. Environmental Science & Technology, 2003. 37(9): p. 1978-1982.
22. Gupta, N., P.K. Roychoudhury, and J.K. Deb, Biotechnology of desulfurization of diesel: prospects and challenges. Applied Microbiology and Biotechnology, 2005. 66(4): p. 356-366.
23. Li, F.L., et al., Deep desulfurization of hydrodesulfurization-treated diesel oil by a facultative thermophilic bacterium Mycobacterium sp. X7B. FEMS Microbiology Letters, 2003. 223(2): p. 301-307.
24. Marcelis, C.L.M., et al., Model description of dibenzothiophene mass transfer in oil/water dispersions with respect to biodesulfurization. Biochemical Engineering Journal, 2003. 16(3): p. 253-264.
25. Olmo, C.H.d., et al., Production of a Rhodococcus erythropolis IGTS8 biocatalyst for DBT biodesulfurization: influence of operational conditions. Biochemical Engineering Journal, 2005. 22(3): p. 229-237.
26. Konishi, M., et al., The separation of oil from an oil-water-bacteria mixture using a hydrophobic tubular membrane. Biochemical Engineering Journal, 2005. 24(1): p. 49-54.
Chapter 7 References
136
27. Velu, S., X.L. Ma, and C.S. Song, Selective adsorption for removing sulfur from jet fuel over zeolite-based adsorbents. Industrial & Engineering Chemistry Research, 2003. 42(21): p. 5293-5304.
28. Hernandez-Maldonado, A.J., et al., New sorbents for desulfurization of diesel fuels via pi complexation: Layered beds and regeneration. Industrial & Engineering Chemistry Research, 2004. 43(3): p. 769-776.
29. Shiraishi, Y., A. Yamada, and T. Hirai, Desulfurization and denitrogenation of light oils by methyl viologen-modified aluminosilicate adsorbent. Energy & Fuels, 2004. 18(5): p. 1400-1404.
30. Haji, S. and C. Erkey, Removal of dibenzothiophene from model diesel by adsorption on carbon aerogels for fuel cell applications. Industrial & Engineering Chemistry Research, 2003. 42(26): p. 6933-6937.
31. Rosso, I., et al., Zinc oxide sorbents for the removal of hydrogen sulfide from Syngas. Industrial & Engineering Chemistry Research, 2003. 42(8): p. 1688-1697.
32. Ma, X., L. Sun, and C. Song, A new approach to deep desulfurization of gasoline, diesel fuel and jet fuel by selective adsorption for ultra-clean fuels and for fuel cell applications. Catalysis Today, 2002. 77(1-2): p. 107-116.
33. Fukunaga, T., et al., Development of kerosene fuel processing system for PEFC. Catalysis Today, 2003. 84(3-4): p. 197-200.
34. Mikhail, S., T. Zaki, and L. Khalil, Desulfurization by an economical adsorption technique. Applied Catalysis A: General, 2002. 227(1-2): p. 265-278.
35. Zainudin, N.F., et al., Study of adsorbent prepared from oil palm ash (OPA) for flue gas desulfurization. Separation and Purification Technology, 2005. In Press, Corrected Proof.
36. Jeevanandam, P., K.J. Klabunde, and S.H. Tetzler, Adsorption of thiophenes out of hydrocarbons using metal impregnated nanocrystalline aluminum oxide. Microporous and Mesoporous Materials, 2005. 79(1-3): p. 101-110.
37. Levy, R.E., et al. Unipure's Oxidative Desulfurization Process Creates New Market Opportunities For Supply Of Ultra-low Sulfur Fuels. in AIChE Spring Meeting. 2002. New Orleans.
38. An, W.D., et al., The electrochemical hydrogenation of edible oils in a solid polymer electrolyte reactor. I. Reactor design and operation. Journal of the American Oil Chemists Society, 1998. 75(8): p. 917-925.
39. An, W.D., et al., The electrochemical hydrogenation of edible oils in a solid polymer electrolyte reactor. II. Hydrogenation selectivity studies. Journal of the American Oil Chemists Society, 1999. 76(2): p. 215-222.
Chapter 7 References
137
40. Perry, R.H. and D.W. Green, eds. Perry's Chemical Engineer's Handbook. 7th ed. ed. 1997, McGraw-Hill: New York.
41. Kroschwitz, J. and M. Howe-Grant, eds. Kirk-Othmer Encyclopedia of Chemical Technology. 4th ed. ed. Vol. Vols. 1, 2, 9, 10. 1991, John Wiley and Son: New York.
42. Gerhartz, W. and Y.S. Yamamoto, eds. Ullman's Encyclopedia of Industrial Chemistry. 5th ed. ed. Vol. Vol. A13. 1996, VCH: Weinheim.
44. Haussinger, P., R. Lohmuller, and A. Watson, Hydrogen, in Ullmann's Encyclopedia of Industrial Chemistry. 2003, Wiley-VCH: Weinheim. p. 85-240.
45. Yi, H.M., Dense palladium and perovskite membranes and membrane reactors. Mrs Bulletin, 1999. 24(3): p. 46-49.
46. Itoh, N., W.C. Xu, and K. Haraya, Basic Experimental-Study on Palladium Membrane Reactors. Journal of Membrane Science, 1992. 66(2-3): p. 149-155.
47. Young, S.K. and K.A. Mauritz, Dynamic mechanical analyses of Nafion (R)/organically modified silicate nanocomposites. Journal of Polymer Science Part B-Polymer Physics, 2001. 39(12): p. 1282-1295.
48. Ogumi, Z., K. Nishio, and S. Yoshizawa, Application of the SPE Method to Organic Electrochemistry. 2. Electrochemical Hydrogenation of Olefinic Double-Bonds. Electrochimica Acta, 1981. 26(12): p. 1779-1782.
49. Ogumi, Z., S. Ohashi, and Z. Takehara, Application of the SPE Method to Organic Electrochemistry. 6. Oxidation of Cyclohexanol to Cyclohexanone on Pt-SPE in the Presence of Iodine and Iodide. Electrochimica Acta, 1985. 30(1): p. 121-124.
50. Ogumi, Z., et al., Application of the Solid Polymer Electrolyte (SPE) Method to Organic Electrochemistry. 3. Kolbe Type Reactions on Pt-SPE. Electrochimica Acta, 1983. 28(11): p. 1687-1693.
51. Inaba, M., Z. Ogumi, and Z. Takehara, Application of the Solid Polymer Electrolyte Method to Organic Electrochemistry .14. Effects of Solvents on the Electroreduction of Nitrobenzene on Cu, Pt-Nafion. Journal of the Electrochemical Society, 1993. 140(1): p. 19-22.
52. Kunugi, A., et al., Electroreduction of 2-cyclohexen-1-one on metal-solid polymer electrolyte composite electrodes. Electrochimica Acta, 1998. 44(4): p. 653-657.
Chapter 7 References
138
53. Sarrazin, J. and A. Tallec, Use of Ion-Exchange Membranes in Preparative Organic Electrochemistry - New Processes for Electrolysis. Journal of Electroanalytical Chemistry, 1982. 137(1): p. 183-188.
54. An, W., J.K. Hong, and P.N. Pintauro, Current efficiency for soybean oil hydrogenation in a solid polymer electrolyte reactor. Journal of Applied Electrochemistry, 1998. 28(9): p. 947-954.
55. Cook, R., R.C. MacDuff, and A.F. Sammells, Ambient Temperature Gas Phase CO2 Reduction to Hydrocarbons at Solid Polymer Electrolyte Cells. Journal of the Electrochemical Society, 1988. 135(6): p. 1470-1471.
56. Komatsu, S., et al., Preparation of Cu-Solid Polymer Electrolyte Composite Electrodes and Application to Gas-Phase Electrochemical Reduction of Co2. Electrochimica Acta, 1995. 40(6): p. 745-753.
57. Lamy-Pitara, E., I. Belegridi, and J. Barbier, Liquid-Phase Catalytic-Hydrogenation under Potential Control of the Catalyst. Catalysis Today, 1995. 24(1-2): p. 151-156.
58. Cleghorn, S.J.C. and D. Pletcher, The Mechanism of Electrocatalytic Hydrogenation of Organic-Molecules at Palladium Black Cathodes. Electrochimica Acta, 1993. 38(2-3): p. 425-430.
59. Horanyi, G., In Horiuti's footsteps: links between catalysis and electrocatalysis. Journal of Molecular Catalysis A - Chemical, 2003. 199(1-2): p. 7-17.
60. Adamson, A.W. and A.P. Gast, Physical Chemistry of Surfaces. 6 ed. 1997, New York: John Wiley & Sons, Inc. 784.
61. Lasia, A. and A. Rami, Kinetics of hydrogen evolution on nickel electrodes. Journal of Electroanalytical Chemistry, 1990(294): p. 123-141.
62. Pardillos-Guindet, J., et al., Electrode Potential of a Dispersed Raney-Nickel Electrode During Acetone Hydrogenation - Influence of the Solution and Reaction-Kinetics. Journal of Catalysis, 1995. 155(1): p. 12-20.
63. Lasser, R. and K.H. Klatt, Solubility of Hydrogen Isotopes in Palladium. Physical Review B, 1983. 28(2): p. 748-758.
64. Langmuir, I., The Adsorption of Gases on Plane Surfaces of Glass, Mica, and Platinum. Journal of the American Chemical Society, 1918. 40: p. 1361-1403.
65. Rositani, F., et al., Kinetics of Acetone Hydrogenation over Pt/Al2O3 Catalysts. Journal of Chemical Technology and Biotechnology A - Chemical Technology, 1985. 35(5): p. 234-240.
Chapter 7 References
139
66. Sun, Y., et al., Asymmetric Hydrogenation of Ethyl Pyruvate: Diffusion Effects on Enantioselectivity. Journal of Catalysis, 1996. 161(2): p. 759-765.
67. Singh, U.K. and M.A. Vannice, Kinetics of liquid-phase hydrogenation reactions over supported metal catalysts -- a review. Applied Catalysis A: General, 2001. 213(1): p. 1-24.
68. Hernandez-Maldonado, A.J. and R.T. Yang, Desulfurization of liquid fuels by adsorption via pi complexation with Cu(I)-Y and Ag-Y zeolites. Industrial & Engineering Chemistry Research, 2003. 42(1): p. 123-129.
69. Popp, F.D. and H.P. Schultz, Electrolytic Reduction of Organic Compounds. Chemical Reviews, 1962. 62(1): p. 19-&.
70. Sharipov, A.K., Mercaptans from Gas Condensate and Crude Oils. Chemistry and Technology of Fuels and Oils, 2002. 38(4): p. 280-285.
71. Satterfield, C.N., Heterogeneous Catalysis in Industrial Practice. 2nd ed. ed. 1991, New York: McGraw-Hill. 554 pages.
72. Control of Air Pollution From New Motor Vehicles: Tier 2 Motor Vehicle Emissions Standards and Gasoline Sulfur Control Requirements, in Federal Register. 2000. p. p. 6697-6870.
73. Naim, M., et al., Inhibition by Thiol Compounds of Off-Flavor Formation in Stored Orange Juice .2. Effect of L-Cysteine and N-Acetyl-L-Cysteine on P-Vinylguaiacol Formation. Journal of Agricultural and Food Chemistry, 1993. 41(9): p. 1359-1361.
74. Mottram, D.S., C. Szauman-Szumski, and A. Dodson, Interaction of thiol and disulfide flavor compounds with food components. Journal of Agricultural and Food Chemistry, 1996. 44(8): p. 2349-2351.
75. Herradura, P.S., K.A. Pendola, and R.K. Guy, Copper-mediated cross-coupling of aryl boronic acids and alkyl thiols. Organic Letters, 2000. 2(14): p. 2019-2022.
76. Reid, E.E., Organic Chemistry of Bivalent Sulfur. Vol. Vol. 1. 1958, New York: Chemical Publishing Co. 539 pages.
77. Tiers, G.V.D. and M.L. Brostrom, Lead mercaptides: materials useful as powder secondary standards and internal references for calibration of X-ray diffractometers at small and medium angles. Journal of Applied Crystallography, 2000. 33(3, Pt. 2): p. 915-920.
78. Borgstrom, P., L.M. Ellis, and E.E. Reid, Preparation, Properties, and Reactions of Lead Mercaptides. Journal of the American Chemical Society, 1929. 51: p. 3649-51.
Chapter 7 References
140
79. Chen, S.W., L.A. Truax, and J.M. Sommers, Alkanethiolate-protected PbS nanoclusters: Synthesis, spectroscopic and electrochemical studies. Chemistry of Materials, 2000. 12(12): p. 3864-3870.
80. Liu, G., et al., Chemistry of sulfur-containing molecules on Au(111): thiophene, sulfur dioxide, and methanethiol adsorption. Surface Science, 2002. 505(1-3): p. 295-307.
81. Benziger, J.B. and R.E. Preston, Organosulfur Chemistry on W(211) Surfaces. 1. A Comparison of Methanethiol and Methanol. Journal of Physical Chemistry, 1985. 89(23): p. 5002-5010.
82. Kuhnle, A., et al., Adsorption of dodecanethiol on Cu(110): Structural ordering upon thiolate formation. Langmuir, 2002. 18(14): p. 5558-5565.
83. Ma, F.Y. and R.B. Lennox, Potential-assisted deposition of alkanethiols on Au: Controlled preparation of single- and mixed-component SAMs. Langmuir, 2000. 16(15): p. 6188-6190.
84. Widrig, C.A., C. Chung, and M.D. Porter, The Electrochemical Desorption of N-Alkanethiol Monolayers from Polycrystalline Au and Ag Electrodes. Journal of Electroanalytical Chemistry, 1991. 310(1-2): p. 335-359.
85. Liu, G., et al., Adsorption of methanethiol on stoichiometric and defective TiO2(110) surfaces: A combined experimental and theoretical study. Journal of Physical Chemistry B, 2002. 106(38): p. 9883-9891.
86. Cheng, L.C., et al., Coadsorption of ethanethiol with sulfur, oxygen, and water on the Fe(100) surface. Langmuir, 1996. 12(2): p. 392-401.
87. Meagher, K.K., et al., Interaction of neopentyl thiol with clean and oxygen-modified Fe(100) surfaces. Journal of Physical Chemistry B, 2000. 104(14): p. 3320-3326.
88. Oae, S., ed. Organic Chemistry of Sulfur. 1st ed. ed. 1977, Plenum Press: New York. 713 pages.
90. Bensebaa, F., et al., XPS studies of metal-sulfur bonds in metal-alkanethiolate materials. Surface Science, 1998. 405: p. L472-476.
91. Souza, A.G.d., et al., Standard enthalpy of formation of di-isobutyldithiocarbamate complexes of zinc-group elements and mean M-S bond-dissociation enthalpies. Thermochimica Acta, 1994. 231: p. 31-38.
Chapter 7 References
141
92. Ribeiro da Silva, M.A.V., A.M.M.V. Reis, and M.M.R. Ferreira da Silva, Enthalpies of formation of crystalline dimethyldithiocarbamate complexes of chromium(III) and iron(III): the estimation of the mean M-S bond-dissociation enthalpies. Journal of Chemical Thermodynamics, 2000. 32: p. 1319-1325.
93. Ribeiro da Silva, M.A.V., A.M.M.V. Reis, and R.I.M.C.P. Faria, Standard enthalpies of formation of crystalline dimethyl ammoniumdimethyldithiocarbamate and dimethyldithiocarbamate complexes of copper(II) and nickel(II). The mean Cu-S and Ni-S bond-dissociation enthalpies. Journal of Chemical Thermodynamics, 1995. 27: p. 1365-1372.
94. Lide, D.R., ed. CRC Handbook of Chemistry and Physics. 81st Ed. ed. 2000, CRC Press: Boca Raton.
95. Barin, I., Thermochemical Data of Pure Substances, Parts I and II. 1993, Weinheim: VCH.
96. Ellis, E.S., et al., Meeting the demands of low sulphur gasoline. Petroleum Technology Quarterly, 2002. Spring.
97. Ondrey, G., This process makes low-sulfur gasoline without additional hydrotreating. Chemical Engineering, 2003. December.
98. Hernandez-Maldonado, A.J., S.D. Stamatis, and R.T. Yang, New Sorbents for Desulfurization of Diesel Fuels via Pi-Complexation: Layered Beds and Regeneration. Industrial and Engineering Chemistry Research, 2004. 43: p. 769-776.
99. Shan, G., et al., Separation of Polycyclic Aromatic Compounds from Model Gasoline by Magnetic Alumina Sorbent Based on Pi-Complexation. Industrial and Engineering Chemistry Research, 2004. 43: p. 758-761.
100. Yang, R.T., A. Takahashi, and F.H. Yang, New sorbents for desulfurization of liquid fuels by, pi-complexation. Industrial & Engineering Chemistry Research, 2001. 40(26): p. 6236-6239.
101. Dvorak, J., T. Jirsak, and J.A. Rodriguez, Fundamental studies of desulfurization processes: reaction of methanethiol on ZnO and Cs/ZnO. Surface Science, 2001. 479(1-3): p. 155-168.
102. Nehlsen, J.P., J.B. Benziger, and I.G. Kevrekidis, Removal of Alkanethiols from a Hydrocarbon Mixture by Heterogeneous Reaction with Metal Oxides. Industrial and Engineering Chemistry Research, 2003. 42(26): p. 6919-6923.
103. Wizig, H. and R. Maple, Removal of Sulfur from a Light FCC Gasoline Stream, 1999, Merichem Chemicals & Refinery Services LLC. http://www.merichem.com/Process/TechnicalArticles/SulfurGas1to3.htm.
Chapter 7 References
142
104. McBryde, W.A.E., Petroleum Deodorized: Early Canadian History of the 'Doctor Sweetening' Process. Annals of Science, 1991. 48: p. 103-111.
105. Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements, in Federal Register. 2001. p. 5001-5051.
106. Dishun, Z., et al., A review of desulfurization of light oil based on selective oxidation. Chemistry Journal on the Internet, 2004. 6(3): p. 17.
107. Song, C., An overview of new approaches to deep desulfurization for ultra-clean gasoline, diesel fuel and jet fuel. Catalysis Today, 2003. 86: p. 211-263.
108. Sweed, N.H. and E.S. Ellis, Meeting the Low Sulfur Mogas Challenge. Petroleum Technology Quarterly, 2001(Autumn).
109. Ford, J.F., T.A. Rayne, and D.G. Adlington, Desulfurization of Hydrocarbons Using Oxidative and Hydro-treatments, US Patent 3,341,448. British Petroleum, Inc., 1967.
110. Babich, I.V. and J.A. Moulijn, Science and technology of novel processes for deep desulfurization of oil refinery streams: a review. Fuel, 2003. 82: p. 607-631.
111. Murata, S., et al., A Novel Oxidative Desulfurization System for Diesel Fuels with Molecular Oxygen in the Presence of Cobalt Catalysts and Aldehydes. Energy & Fuels, 2004. 18: p. 116-121.
112. Tam, P.S., J.R. Kittrell, and J.W. Eldridge, Desulfurization of Fuel-Oil by Oxidation and Extraction .1. Enhancement of Extraction Oil Yield. Industrial & Engineering Chemistry Research, 1990. 29(3): p. 321-324.
113. Shiraishi, Y. and T. Hirai, Desulfurization of vacuum gas oil based on chemical oxidation followed by liquid-liquid extraction. Energy & Fuels, 2004. 18(1): p. 37-40.
115. Collins, F.M., A.R. Lucy, and C. Sharp, Oxidative desulphurisation of oils via hydrogen peroxide and heteropolyanion catalysis. Journal of Molecular Catalysis A-Chemical, 1997. 117(1-3): p. 397-403.
116. Yazu, K., et al., Oxidation of dibenzothiophenes in an organic biphasic system and its application to oxidative desulfurization of light oil. Energy & Fuels, 2001. 15(6): p. 1535-1536.
117. Greenwood, N.N. and A. Earnshaw, eds. Chemistry of the Elements. 2nd ed. 1997, Elsevier. 1341.
Chapter 7 References
143
118. Zhang, Q.L., et al., Reactions between hydrogen sulfide and sulfuric acid: A novel process for sulfur removal and recovery. Industrial & Engineering Chemistry Research, 2000. 39(7): p. 2505-2509.
119. Wang, H., I.G. Dalla Lana, and K.T. Chuang, Kinetics and mechanism of oxidation of hydrogen sulfide by concentrated sulfuric acid. Industrial & Engineering Chemistry Research, 2002. 41(26): p. 6656-6662.
120. Dean, J.A., ed. Lange's Handbook of Chemistry. 15th ed. 1999, McGraw-Hill: New York. 1521.
121. Washburn, E.W., ed. International Critical Tables of Numerical Data, Physics, Chemistry and Technology. 1st Electronic ed. 2003, Knovel. 3414.
122. Muller, H., Sulfur Dioxide, in Ullmann's Encyclopedia of Industrial Chemistry. 2003, Wiley-VCH: Weinheim. p. 659-704.
123. Meadow, J.R. and R.H. Graves, Desulfurization of Heptane Solutions Using Nitrogen Dioxide-Sulfuric Acid Mixtures. Industrial & Engineering Chemistry, 1953. 45(9): p. 2113-2117.
124. Meadow, J.R. and T.A. White, Desulfurization of Heptane Solutions of Organic Sulfur Compounds. Industrial and Engineering Chemistry, 1950. 42(5): p. 925-929.
125. Wood, A.E., A. Lowy, and W.F. Faragher, Action of Petroleum-Refining Agents on Pure Organic Sulfur Compounds Dissolved in Naphtha. Industrial & Engineering Chemistry, 1924. 16(11): p. 1116-1120.
126. Yaws, C.L., ed. Yaws' Handbook of Thermodynamic and Physical Properties of Chemical Compounds. 2003, Knovel.
Acknowledgements
144
Acknowledgements
The work described here represents both a success and a failure. It was a failure
because the original intent, to use membrane reactors to perform some interesting
reactions, didn’t work at all. In fact, few of the ideas that I thought would certainly work
did. Fortunately, careful observation and some curiosity led me to find two unique and
interesting processes that wouldn’t have been discovered were it not for all of the other
failures. I guess the adage is true: we do learn more from our mistakes than our successes.
So goes academic research.
Research, I have come to find, is like a mule. When it moves along, it is capable of
reshaping the world. But most of the time it just sits there, no matter how hard you pull at
the reigns. Persistence is the key. The researcher must be more stubborn than the mule,
constantly coaxing it down the path. It is not always easy to remain so determined. And
so I want to thank all of those who encouraged me and helped me pull this mule.
I want to first thank my advisors, Professors Benziger and Kevrekidis, for allowing
this research to veer off course from time to time. It was largely due to their flexibility
that these processes were discovered. You both have provided excellent guidance and
taught me much about how to think, evaluate, and communicate.
I also want to thank all of my labmates and fellow students. You know perhaps better
than anyone else how difficult it can be to make things work. I want to thank Sonia, Dom,
and Carlos for their friendship and advice during the “early years.” Thank you Barclay,
Acknowledgements
145
Warren, and Joanne for keeping me sane and listening to my unsolicited advice. Thank
you to all of the many REUs and seniors who worked in our lab, especially Joel, for
providing a change of pace and teaching all of us a few things. I thank the “chemistry
people,” Kev, Chris, Jonathan, Paul, Brent, Tao, Lakshmi, and of course Prof. Bocarsly,
for answering my many questions. And thank you to my fellow graduate students for
making my time here an experience I will always cherish.
Of course, none of this would have been possible without the love and support from
my family. To my wife, Amy, thank you for always being there, and being patient with
me. You always believed, even when I didn’t. I also owe the Di Stasio family a great deal
of thanks, and probably a lot more, for their unending support and encouragement. I also
want to thank my family, Mom, Dad, Karen, Hollis, and Katie for convincing me that this
is all worthwhile.
Many people have shaped my time here at Princeton, and I am grateful for them all.
After all, there is no point to learning if we can’t first enjoy life. With those thanks said,
now let’s get on with the science, for that is why we are all here.